In plain words
An interesting failure is one the player can read. They know why it happened, they can see it was their own doing, and they walk away with an idea for next time.
See it
In Jenga, the tower falls because you pulled from a wobbly spot, and everyone saw it coming two moves ago. You know exactly what to try differently, and the collapse is the best laugh of the game.
Why designers care
Players spend most of their time failing at something, so the quality of a game depends on the quality of its failures. Make causes visible, keep each try cheap, add some variety and humor, and failure becomes fuel.
Common mix-up
The bad kinds: failure with no visible cause, and failure where the player knows what to do but can’t physically do it, again and again.